Authors: Research Scholar Pushpa Prajapati, Assistant Professor Dr. Shweta Bhatia
Abstract: The pandemic induced disruption produced a fragmented body of knowledge spanning from human resource management, organisational behaviour, information technology, cybersecurity to the latest integration of artificial intelligence (AI) into workplaces around the remote-hybrid work shift. Initially, quantitative bibliometric analysis was used to map the performance of existing literature by means of Biblioshiny interface of Bibliometrix R package followed by a computational thematic analysis. A peer-reviewed Scopus corpus of 453 articles was funnelled down using PRISMA inclusion-exclusion approach. The performance analysis showed a steep growth in publication during the peak of pandemic transitions in 2023 and 2024; articles were scattered across journals while older articles gained more citations for being early in the lane. The complementary thematic analysis used the title-and-abstract Scopus metadata using topic modelling technique of BERTopic that combined document embeddings, dimensionality reduction using UMAP and density-based clustering using HDBSCAN. The 23 coherent clusters identified were narrowed down to six broad meta-themes, (1) organizational practices, management, and control (2) employee well-being and psychological outcomes (3) technology, knowledge work, and collaboration, (4) spatial and mobility dimensions, (5) risk, security, and productivity and (6) methodological and review contributions. An intertopic distance map was used to study the weight and relative distance of clusters. The triangulation of thematic layers recovered from full title-and-abstract concatenation against standard bibliometric performance indicators, offers a novel keyword-independent literature mapping method and a conceptual framework to reveal underexplored theoretical gaps that have evolved in isolation from one another. Practically, organisational decision-makers have narrow economic evidence for hybrid work rather than employee-experience effects, underscoring the need for empirical future research with unified study design, ideally replicated with transformer-based embeddings and multi-database corpora.
